I build AI agents and automations that take manual work out of real operations. At a behavioral health provider I built the end-to-end revenue cycle management system and shipped LLM agents with RAG over insurer policies, cutting billing follow-up time 70% across $12M+ in claims and dropping claim rejections from 16% to 10%. Before that I was a Lab Engineer and Operations Manager at Amazon, where I ran 105+ benchmarking experiments and led a 100-person fulfillment team. I'm a doctoral candidate at Purdue researching AI adoption, and I do my best work sitting with a team, mapping how the work actually gets done, and shipping a tool they keep using.